Output Parsers
Output parsers transform raw LLM output into structured data. Every parser in Synaptic implements the Runnable trait, so they compose naturally with prompt templates, chat models, and other runnables using the LCEL pipe operator (|).
Available Parsers
| Parser | Input | Output | Description |
|---|---|---|---|
StrOutputParser | Message | String | Extracts the text content from a message |
JsonOutputParser | String | serde_json::Value | Parses a string as JSON |
StructuredOutputParser<T> | String | T | Deserializes JSON into a typed struct |
ListOutputParser | String | Vec<String> | Splits by a configurable separator |
EnumOutputParser | String | String | Validates against a list of allowed values |
BooleanOutputParser | String | bool | Parses yes/no/true/false strings |
MarkdownListOutputParser | String | Vec<String> | Parses markdown bullet lists |
NumberedListOutputParser | String | Vec<String> | Parses numbered lists |
XmlOutputParser | String | XmlElement | Parses XML into a tree structure |
All parsers also implement the FormatInstructions trait, which provides a get_format_instructions() method. You can include these instructions in your prompt to guide the LLM toward producing output in the expected format.
Quick Example
use synaptic::parsers::StrOutputParser;
use synaptic::runnables::Runnable;
use synaptic::core::{Message, RunnableConfig};
let parser = StrOutputParser;
let config = RunnableConfig::default();
let result = parser.invoke(Message::ai("Hello world"), &config).await?;
assert_eq!(result, "Hello world");Sub-Pages
- Basic Parsers -- StrOutputParser, JsonOutputParser, ListOutputParser
- Structured Parser -- deserialize JSON into typed Rust structs
- Enum Parser -- validate output against a fixed set of values